Transcranial magnetic stimulation (TMS) is a non-invasive method for probing and modulating human brain function. It is approved for the treatment of depression and pre-surgical cortical mapping; it also shows promise in other neurological and psychiatric disorders. Exactly what TMS does to neuronal activity, however, remains unknown. This gap in our knowledge precludes us from biologically-based, rational design of TMS protocols. To fill this gap, we need a better mechanistic understanding of the effect of TMS on cerebral neurons and a database of dose-response curves that describe how the selection of TMS parameters (the dose) relates to changes of neuronal activity (the response). Our project aims to contribute such mechanistic insight and empirical data. Our interdisciplinary team has developed a novel repertoire of tools and techniques that permit us to manipulate the TMS stimulus parameters, model the resulting electromagnetic fields and neuronal responses, and record from cerebral neurons while TMS is applied. In our first set of experiments (Aim 1), we will vary the temporal parameters of TMS. Using a custom TMS pulse generator, we will systematically change what the individual pulses look like (the pulse waveform) and how they are applied sequentially (the pulse train). Concomitant recordings in the zone of stimulation will determine how the various parameters modulate the firing rates of axons, excitatory neurons, and inhibitory neurons. Second (Aim 2), we will vary the spatial parameters of TMS using various coil locations and types of stimulation coils, including macaque-scaled approximations to conventional figure-8 coils as well as less focal coils recently approved for depression treatment (H coils). With simultaneous targeted recordings in the brain we will map the neural response in various cortical regions. In parallel to these empirical studies, we will construct individual, realistic, MRI-based head models coupled with neural response models to simulate, respectively, the electric field spatial distribution and the resultin response of various neuron types. The simulations will both guide and be informed by the empirical neural recordings, enhancing our understating of the mechanisms of TMS and providing a novel simulation tool that could inform TMS dosage. The end result of this project will be to discover how TMS influences the brain at the level of single neurons and simple circuits. The outcome should be transformative in helping researchers and clinicians to navigate the vast parametric space of TMS so that it may be used more effectively as a probe in neuroscience, and as a clinical treatment.